Exploratory Data Analysis on UKRAINE RUSSIA War (2022)

The EDA on the war will use two datasets.
Important: Each new record is accumulated data from previous days.
Important. Data will be updated daily

We will talk about the following:

  1. Tracking
  2. Personnel
  3. Prisoner of War
  4. Armored Personnel Carrier
  5. Multiple Rocket Launcher
  6. Aircraft
  7. Anti-aircraft warfare
  8. Drone
  9. Field Artillery
  10. Fuel Tank
  11. Helicopter
  12. Military Auto
  13. Naval Ship
  14. Tank

Source of data:source

Acronyms

  1. POW - Prisoner of War
  2. MRL - Multiple Rocket Launcher
  3. APC - Armored Personnel Carrier
  4. SRBM - Short-range ballistic missile
  5. drones: UAV - Unmanned Aerial Vehicle; RPA - Remotely Piloted Vehicle.

Import the libraries

Data Preparation and Cleaning

Dataframe for losses on equipments

Dataframe for losses on personnel and POW

Merge both the dataframes



We choose to replace null values with zero as data has not been provided for "NaN" values.

Note: The dataset conatins a combination of equipments lost for both entities.

Before carrying out any analysis let us cross check our data with real estimates of Russian firepower in comparison with columns in our dataset.

Personnel / Manpower:
Total Military Personnel(est.) : 1,350,000
Total Military Personnel can be broken down into three categories:

  1. Active Personnel: 850,000
  2. Reserve Personnel: 250,000
  3. Paramilitary: 250,000

Airpower:
Total Aircraft strength:4,173

  1. Interceptors (Combat Aircraft): 772
  2. Helicopters: 1,543
  3. Dedicated Attack (Total Attack aircraft): 739
  4. Transports (Fixed wing transport): 445
  5. Trainers (Military trainer aircraft): 522
  6. Special Mission (Specil mission aircraft): 132
  7. Tanker fleet (Tanker fleet aircraft): 20
  8. Attack Helicopters (Military attack helicopters): 544

Land Forces:

  1. Tanks : 12,420
  2. Armored Vehicles: 30,122
  3. Self-propelled artillery: 6,574
  4. Towed Artillery: 7,571
  5. Rocket Projectors: 3,391

Naval Forces:

  1. Total Assets: 605
  2. Aircraft Carriers: 1
  3. Destroyers: 15
  4. Frigates: 11
  5. Corvettes: 86
  6. Submarines: 70
  7. Patrol Vessels: 59
  8. Mine Warfare: 49 source

NOTE: COMPARE THIS DATA WITH THE TOTAL LOSSES RECORDED IN OUR DATASET

Our losses of helicopters in is far more than the inventory of helicopters in Russia.
The Ukrainian Air Force has: 69 fighters 29 dedicated attack aircraft 32 transports 71 trainers 5 aircraft designated as special mission 112 helicopters source
Combined inventory of helicopters is still less than the losses we have recorded.
The same can be said about the personnel lost.

This is because every row in our dataframe returns the total number of a particular equipment and personnel loss recorded up to that date

SO WHAT TO DO NOW????

IN ORDER TO PROCEED our first step will be to create a revised dataframe that will contain the equipment and personnel lossesfor every 24hrs.

Check for outliers

Create a distplot for equipments, personnel and POW



We will avoid treating the outliers as any manipultaion of outliers will lead to a significant loss of data for different dates

In our case we need numerical values to accurately analyze the trend for different types of equipments, personnel and POW

Exploratory Analysis and Visualization

On which date the equipments lost was the highest and the lowest?



So far the maximum amount of losses have been recorded on 2022-02-25 and the minimum on 2022-02-27. This will be subjected changes that will be dependent on data update.

Which equipment had the maximum and minimum amount of losses?

Equipment with least losses



Equipment with min. losses: Mobile Short Range Ballistic Missile (SRBM) System

Equipment with most losses



Equipment with max. losses : APC

Plot for all equipments lost over time

Trend for each equipment lost


We create a function that comprises of three different operations:

  1. The line plot is to identify the trend for every type of loss for every single day of the ongoing war.
  2. The bar plot is to display the percentage drop or rise of losses from their previous day.
  3. We also return a dataframe that show us the amount of equipment lost and their respective percentage rise or drop in comparison to the previous day of the war.

AIRCRAFTS

HELICOPTER

TANK

APC

FIELD ARTILLERY

MRL

MILITARY AUTO

FUEL TANK

DRONES

NAVAL SHIP

ANTI AIRCRAFT WARFARE

SPECIAL EQUIPMENT

MOBILE SRBM SYSTEM

PERSONNEL

PRISONER OF WAR

Total Personnel Losses and POW

Aerial eqiupments during war

Create a database for the aerial equipments lost.



We will create a new dataframe from our dataframe of losses for aerial equipments.
This dataframe will help us see the contribution of different aerial equipments (percentage) towards the total of aerial equipments lost

Ground equipments during war

Create a database for the ground equipments lost.



We will create a new dataframe from our dataframe of losses for ground equipments.
This dataframe will help us see the contribution of different ground equipments (percentage) towards the total of ground equipments lost

Create a database for the naval equipments lost.

The war signals towards a heavy loss of ground equipments. Whereas the use of naval equipments is almost null. Let us calculate the total for each category.

We can see that the amount of losses of ground equipment is far more than the losses in other categories till now.